I saw this tweet by Sam Lessin on why there won’t be app interfaces in the future.

I disagree.

Chat interfaces first became popular in 2015 and everyone started building chat workflows. Then everyone was under the impression that Alexa (and other voice interfaces) would do everyone’s job for them.

Reality: You can’t build complex workflows over chat or voice.

Take food ordering. A single use case. Not a lot of dependencies. Swiggy already has all of your data in place.

Now let’s say Swiggy doesn’t have an app. It’s a chat interface. Now walk through your typical food order journey. Think about how you went about ordering to eat yesterday. Now assume that you had to do it through voice or chat as opposed to the current app that Swiggy has. Which one is faster/easier?

Now it’s no longer a hypothetical glorious future use case. It’s something you can think of now.

Do this simple exercise if you are a designer or a PM.

Yes, you can eliminate steps in the current process. And Siri might actually help you with your frequent use cases.

Even Gojek has a one-tap feature in our product for users for whom we can predict their destination, drop-off, have wallet information. So with a single tap, you can bypass all the steps that you have to do today on Uber / Ola, which does not have that feature.

But you don’t build a ride-hailing product on the basis of one happy use case.

Even for people using one-tap, you need to figure out how this will convert compared to the standard booking flow. More importantly, how many people have added all the info so that you can create that magical one-tap experience.

As someone who has been building marketplace products for the last 6 years, I can tell you all the complexities of a car marketplace, a food marketplace, a grocery marketplace or a logistics marketplace.

But it is going to take a lot more than a Twitter thread. The key is always in the details.

Context is important. The implementation details matter.

Go back to all the thought leadership posts on chat is the new interface from a decade ago. Do the same for voice. Realise how they missed key problems. They will still be there.

Will the new Apple Intelligence workflows be well received? Sure. Are they going to be as disruptive as people are making them out to be? Whenever people make very bold predictions about the future, I ignore them.

I still have the memory of Chamath’s screaming that Facebook is dead. Last I checked it is doing quite fine. I remember his blockchain will be the death of Visa and Mastercard comment. The last time I checked, the products and share prices of these companies are doing quite well.

Thinking from first principles. And when people say X is dead and Y is the new way, you don’t have to panic.